Non-contact respiration detection method based on computer vision

Abstract

The invention relates to a computer vision-based non-contact breath detection method, which comprises the following steps of: acquiring a video image, dividing an ROI (Region of Interest) containing the chest and abdomen of a human body in the video image, and calculating the area S1 of the ROI; semantic segmentation is carried out on the image in the ROI, and the image area S2 contained in the chest and abdomen contour of the human body is calculated. Firstly, an OTSU algorithm is used for carrying out binarization processing on an image; then carrying out opening operation, i.e., firstly corroding and then expanding, so as to eliminate a white region with a smaller area; and carrying out closed operation on the processed image, namely, firstly expanding and then corroding, so as to eliminate a small black area. After the above processing is completed, if the number of the connected domain is 1, calculating the area S2 of the connected domain, namely the area of the speech region contained in the chest and abdomen of the human body; if the number of the connected domains is larger than or equal to 2, the maximum connected domain area is calculated and recorded as S2; and then calculating the ratio K of the S2 to the area S1 of the ROI, namely K = S2 / S1.

Description

technical field [0001] The invention relates to a breathing detection method, in particular to a non-contact method for detecting human breathing based on computer vision, which is suitable for special groups such as newborns, patients with extensive burns, and paralyzed patients. Background technique [0002] Respiration is one of the most important vital signs of the human body. The clinical detection of breathing frequency and abnormal breathing state is mainly based on the contact method, that is, the ECG electrodes are attached to the surface of the patient's chest, and the measured leads are measured when the breathing movement is measured by the monitor. The change in impedance between them is used to convert the respiratory rate. [0003] This contact detection method has certain shortcomings, especially for special populations, such as the underdeveloped skin of newborns, the loose connection between the epidermis and the dermis, and attaching electrodes to the skin...

AI Technical Summary

Problems solved by technology

The defect of this method lies in the following two points: 1. It is easily affected by changes in the brightness of the surrounding environment, such as the brightness changes caused by the walking of the nursing staff in the neonatal ward and the brightness interference caused by the blue light treatment of the surrounding newborns

2. The aliasing problem of the heartbeat signal and the breathing signal. This method is to distinguish the two by setting different cut-off frequencies. This method is relatively limited, because the breathing frequency and heart rate of special groups of people in real life are very different from normal people. For example, the respiratory rate of adults is 12-20 times / min, and the heart rate is 60-100 times / min, while the respiratory rate of newborns is significantly higher than that of adults, which is 40-60 times / min. Abnormal, if the cut-off frequency is set, the two signals cannot be correctly screened and distinguished, and it is impossible to effectively judge whether it is abnormal

The disadvantage of this method is that it is not suitable for special populations to use double-sided tape to attach markers on the skin surface, especially newborns and patients with extensive burns

Images